Solution-processed amorphous metal oxide semiconductors (AOSs) are promising candidates for printed electronics. However, process durability and bias stress instability issues still hinder their practical applications. Here, a poly(methyl methacrylate)/parylene/AlOx hybrid passivation approach was developed for AOS thin-film transistors (TFTs) to overcome these challenges. Notably, AlOx was successfully formed without degrading the AOS TFTs owing to the polymer buffer layers. The hybrid passivation approach ensured the satisfactory stabilization of the TFTs under bias stress owing to the high isolation effect, which could prevent the penetration of environmental molecules. This passivation method can facilitate the application of solution-processed AOSs in integrated circuits.

Utilizing printed electronics is a promising way to realize the Internet of Things (IoT) society because of its high throughput. In particular, thin-film transistors (TFTs) based on printable semiconductors are regarded as key devices toward developing trillions of sensors and radiofrequency identification tags, thereby realizing information exchange between different objects. Among different printable materials, solution-processed amorphous metal oxide semiconductors (AOSs) have been considered as potential candidates for printing TFTs owing to their high carrier mobility and excellent uniformity.1 However, their sensitivity toward surrounding molecules and environments, such as variable mobility, as well as threshold voltage and bias stress instability, makes AOS-based TFTs unstable during fabrication processes and practical operations.2–4 Such problems are more pronounced for TFTs based on solution-processed AOSs because of their lower active layer quality.5–7 Specifically, different from vacuum-processed AOSs, solution-processed AOS films comprise numerous voids and organic residues, and they generally have low densities and are thin.8–11 Hence, device stability should be investigated to print advanced electronics toward the realization of the IoT society with solution-processed AOSs.

Passivation is a simple and effective way to improve device stability by curbing the penetration of surrounding molecules, including oxygen and water, to the channel layers.12 Some representative encapsulants, including insulating inorganics, such as SiOx,13,14 SiOx/SiNx,15 AlOx,16–19 and YOx,20 exhibit low gas and water permeability. However, TFT performance degradation has been observed after their deposition processes, such as plasma-enhanced chemical vapor deposition (CVD), atomic layer deposition (ALD), and sputtering.17,21,22 Another encapsulant is insulating polymers, such as CYTOP,21,23 parylene,24 and some reactive resins,25–27 which are compatible with mild processing conditions, and thus generate inconsiderable process-induced damage. However, such polymers exhibit insufficient protection abilities owing to their high gas and water penetration rates compared with their inorganic counterparts.22 Some hybrid systems, such as using a bilayer structure22,28,29 and organic–inorganic hybrid polymers,30–32 have been developed to take advantage of both materials. However, many passivation techniques used on vacuum-processed AOSs could not be directly adopted for solution-processed AOSs because of their high environmental sensitivity. Therefore, it is necessary to develop harmless and effective passivation methods suitable for solution-processed AOS TFTs.

In this study, a poly(methyl methacrylate) (PMMA)/parylene/AlOx hybrid passivation technique was developed to simultaneously obtain strong protection without performance degradation. With this hybrid structure, a thin PMMA/parylene bilayer acted as a buffer from damage resulting from the ALD process for AlOx, and the resultant AlOx had a strong isolation effect. This facile and highly effective passivation technology will be suitable for facilitating the realization of printed electronics in the future.

Indium zinc oxide (IZO) was used as the active layer in this study. An IZO precursor was prepared by mixing 0.1M In(NO3)3 · xH2O (Aldrich) and Zn(NO3)2 · 6H2O (Aldrich) solutions (2-methoxyethanol as a solvent) at an In/Zn mole ratio of 3:2. A bottom-gate, top-contact IZO TFT with the hybrid passivation layer is illustrated in Fig. 1(a). Gate electrodes (Cr/Au/Cr, 5/25/5 nm) were fabricated by means of a conventional lift-off process on a glass substrate. A gate dielectric layer (AlOx, 75 nm) was formed via ALD. Thereafter, the IZO precursor was spin-coated at 5000 rpm for 30 s and annealed at 370 °C for 1 h. Subsequently, photolithography was performed using a photosensitive dielectric material (PDM, Taiyo Ink Mgf. Co., Ltd.)33 to pattern the IZO active layer, as well as source and drain (S/D) electrodes (Al, 40 nm). Furthermore, the IZO was etched with oxalic acid (0.075 mg/100 ml), and the S/D electrodes were formed via a lift-off process. The details of fabrication can be found in our previous paper.34 For passivation, PMMA (Mw: 120 000) was formed using the spin-coating technique with a butyl acetate solution (5 mg/100 ml), followed by annealing at 150 °C for 1 h. Parylene (25 nm) and AlOx (40 nm) were deposited via CVD and ALD, respectively. The ALD process was carried out with maintaining the substrate temperature at 110 °C. After both CVD and ALD, post-annealing process was not performed. All electrical measurements were conducted under ambient and dark conditions using a semiconductor parameter analyzer (Keithley 4200-SCS). The thickness of the IZO active layer and PMMA was estimated by means of x-ray reflectometry (XRR, Rigaku SmartLab with CuKα irradiation, λ = 1.54056 Å). Atomic force microscopy (AFM) was conducted using a Shimadzu SPM-9700HT.

FIG. 1.

(a) Schematic TFT structure with PMMA/parylene/AlOx hybrid passivation layer. (b) Optical micrographs of a TFT with L/W = 10 µm/200 µm.

FIG. 1.

(a) Schematic TFT structure with PMMA/parylene/AlOx hybrid passivation layer. (b) Optical micrographs of a TFT with L/W = 10 µm/200 µm.

Close modal

Optical micrographs of an as-fabricated IZO TFT with a channel length (L) and width (W) of 10 µm and 200 µm, respectively, are shown in Fig. 1(b). Here, we discuss different passivation structures, including PMMA single layer, parylene single layer, AlOx single layer, PMMA/parylene bilayer, PMMA/AlOx bilayer, and PMMA/parylene/AlOx trilayer. The transfer characteristics were measured before and after passivation to evaluate the process-induced performance degradation. Figure 2(a) summarizes the transfer curves of the TFTs before and after being passivated with PMMA, parylene, and AlOx single layers. Before passivation, the TFT had a saturation mobility (μsat) of 2.3 cm2 V−1 s−1, on/off current ratio (Ion/Ioff) of ∼108, and threshold voltage (Vth) of 2.0 V. Parylene passivation led to a large negative Vth shift, which might be caused by the long-time vacuum state during parylene deposition and/or reactive radical species derived from the parylene source that might react with weakly bonded oxygen atoms. The TFT with AlOx passivation performed similarly, but worse than the parylene-passivated one. This may be due to the harsher processing conditions, including high vacuum together with elevated temperature (110 °C) and/or using water as a reaction species. A similar phenomenon has been reported for vacuum-processed AOSs.17,22 Hence, ALD and parylene-coating processes tend to vary the TFT properties based on solution-processed AOSs. Nonetheless, after PMMA passivation, TFT still exhibited desirable properties, including μsat of 2.0 cm2 V−1 s−1, Ion/Ioff ≈ 108, Vth of 1.7 V, and negligible hysteresis, thereby indicating the suitability of PMMA as a direct encapsulant for the IZO active layer. This result is attributable to the mild fabrication conditions of PMMA, i.e., the spin-coating method under ambient conditions. However, the high moisture and gas penetration rates render PMMA insufficient for passivation.29,35,36 To solve this problem, we used PMMA as a buffer layer and deposited either parylene or AlOx on top of it. The transfer characteristics of the resulting PMMA/parylene and PMMA/AlOx bilayer passivation layers are summarized in Fig. 2(b); for the former, only minimal changes are observed, suggesting that the PMMA layer can effectively protect IZO TFTs during parylene deposition, despite its thickness being only 13 nm [Fig. 2(c)]. Thus, the flat and pinhole-free structure of the PMMA layer is suitable for preventing radical species from penetrating into the IZO layer [Figs. 2(d) and 2(e)]. On the one hand, the PMMA buffer layer was unable to provide sufficient protection during the ALD of AlOx because of the harsher deposition conditions of ALD compared to that of the parylene coating. On the other hand, the PMMA/parylene bilayer structure acted as a strong buffer layer for the TFTs during the ALD of AlOx, whereby evident TFT performance degradation was not observed [Fig. 2(f)].

FIG. 2.

Transfer characteristics of TFTs with and without (a) single layer and (b) bilayer passivation. (c) XRR data of the Si/SiO2 (100 nm)/IZO/PMMA structure. The best fitting result provides 6 nm and 13 nm of IZO and PMMA layers, respectively. AFM images of (d) IZO fabricated on a Si/SiO2 substrate and (e) PMMA formed on top of the IZO. (f) Transfer characteristics of TFTs with and without trilayer passivation. Drain voltage (VD) is 10 V for all transfer curves.

FIG. 2.

Transfer characteristics of TFTs with and without (a) single layer and (b) bilayer passivation. (c) XRR data of the Si/SiO2 (100 nm)/IZO/PMMA structure. The best fitting result provides 6 nm and 13 nm of IZO and PMMA layers, respectively. AFM images of (d) IZO fabricated on a Si/SiO2 substrate and (e) PMMA formed on top of the IZO. (f) Transfer characteristics of TFTs with and without trilayer passivation. Drain voltage (VD) is 10 V for all transfer curves.

Close modal

To further understand the isolation ability of PMMA and PMMA/parylene buffer layers, their waterproof properties were evaluated by immersing the TFTs in deionized water for 15 min, air-dried, and then measured and compared the transfer characteristics of different devices [Fig. 3(a)]. The transfer curves of TFTs with PMMA, PMMA/parylene passivation, and TFT without passivation after water treatment are shown in Fig. 3(b). A TFT without passivation and water treatment was used as a reference. Without passivation, a considerably negative Vth shift and large hysteresis were observed because of the large amount of water existing on the backchannel layer acting as the donor. The large hysteresis might also be generated because of the high-humidity conditions that could cause additional effects, such as peeling of the active layer, thereby resulting in a severe deterioration in TFT performances.5 Conversely, TFTs with passivation layers exhibited a smaller Vth shift and negligible hysteresis. Compared with the PMMA single layers, PMMA/parylene bilayers exhibited stronger protection abilities. This result is consistent with the phenomenon that the PMMA/parylene bilayer has better buffer functions during AlOx deposition than the PMMA single layer because of the low moisture penetration rate of parylene.24 

FIG. 3.

(a) Schematic of water treatment. (b) Transfer characteristics of TFTs before and after water treatment (VD = 10 V).

FIG. 3.

(a) Schematic of water treatment. (b) Transfer characteristics of TFTs before and after water treatment (VD = 10 V).

Close modal
Further advances in the present trilayer passivation approach were studied in terms of bias stress stability at room temperature (25 °C) under ambient conditions. The drain current (ID) was recorded throughout the process, and the ID drop rate (defined in the following equation) after 3700 s was used to compare the stability of the TFTs with different passivation structures. The results are shown in Fig. 4(a).
IDdroprate=1ID(t)ID(0)×100%,
(1)
where ID(0) is the initial drain current and ID(t) is the ID after applying bias stress for the respective time (t). Under constant gate bias, all TFTs exhibited the same tendency of ID dropping gradually, which was probably because the absorbed oxygen could capture electrons from the conductive bond, expressed as O2(g) + 2e(s) = 2O(s).4,37 The TFT without passivation exhibited the poorest bias stress stability in this work with an ID drop rate of 67%. However, polymer passivation layers could suppress the ID drop to a certain extent, reaching 17% and 21% for PMMA- and PMMA/parylene-passivated TFTs, respectively. Notably, the ID drop was suppressed significantly by the PMMA/parylene/AlOx trilayer passivation with an ID drop rate of only 2%. In addition, the transfer characteristics were studied before and after the bias stress to estimate the Vth shift value (ΔVth). Without passivation, ΔVth was estimated to be 2.8 V. In the case of the single PMMA [Fig. 4(c)] and PMMA/parylene bilayer passivation processes [Fig. 4(d)], ΔVth decreased to 0.7 V and 0.8 V, respectively. Significantly, the PMMA/parylene/AlOx trilayer passivation approach suppressed ΔVth quite effectively and showed ΔVth of only 0.2 V [Fig. 4(e)] owing to the high isolation effect of ALD-processed AlOx without degrading the IZO and the PMMA/parylene buffer layers.
FIG. 4.

(a) ID changes during the bias stress test for TFTs with and without passivation structures (gate voltage, VG = 10 V; VD = 10 V). Transfer characteristics before and after bias stress for 3700 s of TFTs (b) without and with (c) single PMMA, (d) PMMA/parylene bilayer, and (e) PMMA/parylene/AlOx trilayer passivation (VD = 10 V).

FIG. 4.

(a) ID changes during the bias stress test for TFTs with and without passivation structures (gate voltage, VG = 10 V; VD = 10 V). Transfer characteristics before and after bias stress for 3700 s of TFTs (b) without and with (c) single PMMA, (d) PMMA/parylene bilayer, and (e) PMMA/parylene/AlOx trilayer passivation (VD = 10 V).

Close modal

In this study, a hybrid trilayer passivation method was developed to realize the high stability of TFTs based on solution-processed IZO. With the help of a thin PMMA buffer layer, process-induced damage during parylene deposition was perfectly precluded. In addition, owing to the strong protection of the PMMA/parylene buffer layer, performance deterioration during AlOx deposition via ALD was effectively suppressed. Owing to the high isolation effect of the hybrid passivation layer, the IZO-based TFTs were significantly stabilized under bias stress in air, whereby the ID drop rate was reduced from 67% in the non-passivated TFT to 2%, while ΔVth decreased from 2.8 V to 0.2 V. Therefore, the present passivation technique provides a near-ideal protection of solution-processed AOS TFTs, and it could improve the future availability of printed electronics for the realization of the IoT society.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
J. W.
Park
,
B. H.
Kang
, and
H. J.
Kim
,
Adv. Funct. Mater.
30
,
1904632
(
2020
).
2.
J.-S.
Park
,
J. K.
Jeong
,
H.-J.
Chung
,
Y.-G.
Mo
, and
H. D.
Kim
,
Appl. Phys. Lett.
92
,
072104
(
2008
).
3.
A.
Suresh
and
J. F.
Muth
,
Appl. Phys. Lett.
92
,
033502
(
2008
).
4.
S.-Y.
Sung
,
J. H.
Choi
,
U. B.
Han
,
K. C.
Lee
,
J.-H.
Lee
,
J.-J.
Kim
,
W.
Lim
,
S. J.
Pearton
,
D. P.
Norton
, and
Y.-W.
Heo
,
Appl. Phys. Lett.
96
,
102107
(
2010
).
5.
S.-U.
Lee
and
J.
Jeong
,
IEEE J. Electron Devices Soc.
7
,
26
(
2019
).
6.
W.
Xu
,
H.
Li
,
J.-B.
Xu
, and
L.
Wang
,
ACS Appl. Mater. Interfaces
10
,
25878
(
2018
).
7.
A. S.
Safaruddin
,
J. P. S.
Bermundo
,
N.
Yoshida
,
T.
Nonaka
,
M. N.
Fujii
,
Y.
Ishikawa
, and
Y.
Uraoka
,
Appl. Phys. Express
12
,
064002
(
2019
).
8.
J.
Socratous
,
K. K.
Banger
,
Y.
Vaynzof
,
A.
Sadhanala
,
A. D.
Brown
,
A.
Sepe
,
U.
Steiner
, and
H.
Sirringhaus
,
Adv. Funct. Mater.
25
,
1873
(
2015
).
9.
Y.
Chen
,
B.
Wang
,
W.
Huang
,
X.
Zhang
,
G.
Wang
,
M. J.
Leonardi
,
Y.
Huang
,
Z.
Lu
,
T. J.
Marks
, and
A.
Facchetti
,
Chem. Mater.
30
,
3323
(
2018
).
10.
J. W.
Hennek
,
J.
Smith
,
A.
Yan
,
M.-G.
Kim
,
W.
Zhao
,
V. P.
Dravid
,
A.
Facchetti
, and
T. J.
Marks
,
J. Am. Chem. Soc.
135
,
10729
(
2013
).
11.
X.
Yu
,
J.
Smith
,
N.
Zhou
,
L.
Zeng
,
P.
Guo
,
Y.
Xia
,
A.
Alvarez
,
S.
Aghion
,
H.
Lin
,
J.
Yu
,
R. P. H.
Chang
,
M. J.
Bedzyk
,
R.
Ferragut
,
T. J.
Marks
, and
A.
Facchetti
,
Proc. Natl. Acad. Sci. U. S. A.
112
,
3217
(
2015
).
12.
L.
Petti
,
N.
Münzenrieder
,
C.
Vogt
,
H.
Faber
,
L.
Büthe
,
G.
Cantarella
,
F.
Bottacchi
,
T. D.
Anthopoulos
, and
G.
Tröster
,
Appl. Phys. Rev.
3
,
021303
(
2016
).
13.
X.
Liu
,
W.
Wu
,
W.
Chen
,
H.
Ning
,
X.
Zhang
,
W.
Yuan
,
M.
Xiong
,
X.
Wang
,
R.
Yao
, and
J.
Peng
,
Materials
11
,
1440
(
2018
).
14.
C.
Dong
,
J.
Shi
,
J.
Wu
,
Y.
Chen
,
D.
Zhou
,
Z.
Hu
,
H.
Xie
,
R.
Zhan
, and
Z.
Zou
,
Mater. Sci. Semicond. Process.
20
,
7
(
2014
).
15.
C.
Zhao
,
T.-C.
Fung
, and
J.
Kanicki
,
Solid-State Electron.
120
,
25
(
2016
).
16.
S.
Yang
,
D.-H.
Cho
,
M. K.
Ryu
,
S.-H. K.
Park
,
C.-S.
Hwang
,
J.
Jang
, and
J. K.
Jeong
,
Appl. Phys. Lett.
96
,
213511
(
2010
).
17.
T.
Mudgal
,
N.
Edwards
,
P.
Ganesh
,
A.
Bharadwaj
,
R. G.
Manley
, and
K. D.
Hirschman
,
ECS Trans.
72
,
67
(
2016
).
18.
S.
Hu
,
K.
Lu
,
H.
Ning
,
Z.
Zheng
,
H.
Zhang
,
Z.
Fang
,
R.
Yao
,
M.
Xu
,
L.
Wang
,
L.
Lan
,
J.
Peng
, and
X.
Lu
,
IEEE Electron Device Lett.
38
,
879
(
2017
).
19.
J.-S.
Seo
,
J.-H.
Jeon
,
Y. H.
Hwang
,
H.
Park
,
M.
Ryu
,
S.-H. K.
Park
, and
B.-S.
Bae
,
Sci. Rep.
3
,
2085
(
2013
).
20.
K.
Nomura
,
T.
Kamiya
, and
H.
Hosono
,
Thin Solid Films
520
,
3778
(
2012
).
21.
S.-H.
Choi
,
J.-H.
Jang
,
J.-J.
Kim
, and
M.-K.
Han
,
IEEE Electron Device Lett.
33
,
381
(
2012
).
22.
S.
Zhan
,
S.
Han
,
S. Y.
Bang
,
B.
Li
,
Y. T.
Chun
,
B.
Hou
, and
J. M.
Kim
,
Phys. Status Solidi A
217
,
1900832
(
2020
).
23.
K.-M.
Jung
,
J.
Oh
,
H. E.
Kim
,
A.
Schuck
,
K.
Kim
,
K.
Park
,
J.-H.
Jeon
,
S.-Y.
Lee
, and
Y.-S.
Kim
,
J. Phys. D: Appl. Phys.
53
,
355107
(
2020
).
24.
A.
Kiazadeh
,
H. L.
Gomes
,
P.
Barquinha
,
J.
Martins
,
A.
Rovisco
,
J. V.
Pinto
,
R.
Martins
, and
E.
Fortunato
,
Appl. Phys. Lett.
109
,
051606
(
2016
).
25.
A.
Olziersky
,
P.
Barquinha
,
A.
Vilà
,
L.
Pereira
,
G.
Gonçalves
,
E.
Fortunato
,
R.
Martins
, and
J. R.
Morante
,
J. Appl. Phys.
108
,
064505
(
2010
).
26.
H.
Lu
,
T.
Liang
, and
S.
Zhang
, in
2016 13th IEEE International Conference on Solid-State and Integrated Circuit Technology (ICSICT)
(
IEEE
,
2016
), p.
638
.
27.
H.
Yoo
,
Y. J.
Tak
,
W.-G.
Kim
,
Y.-g.
Kim
, and
H. J.
Kim
,
J. Mater. Chem. C
6
,
6187
(
2018
).
28.
C.-L.
Fan
,
M.-C.
Shang
,
B.-J.
Li
,
Y.-Z.
Lin
,
S.-J.
Wang
,
W.-D.
Lee
, and
B.-R.
Hung
,
Materials
8
,
1704
(
2015
).
29.
S. K.
Park
,
Y.-H.
Kim
,
H.-S.
Kim
, and
J.-I.
Han
,
Electrochem. Solid-State Lett.
12
,
H256
(
2009
).
30.
J. P.
Bermundo
,
Y.
Ishikawa
,
H.
Yamazaki
,
T.
Nonaka
,
M. N.
Fujii
, and
Y.
Uraoka
,
Appl. Phys. Lett.
107
,
033504
(
2015
).
31.
N.
Yoshida
,
J. P.
Bermundo
,
Y.
Ishikawa
,
T.
Nonaka
,
K.
Taniguchi
, and
Y.
Uraoka
,
Appl. Phys. Lett.
112
,
213503
(
2018
).
32.
A. S.
Safaruddin
,
J. P. S.
Bermundo
,
N.
Yoshida
,
T.
Nonaka
,
M. N.
Fujii
,
Y.
Ishikawa
, and
Y.
Uraoka
,
IEEE Electron Device Lett.
41
,
1372
(
2020
).
33.
X.
Wei
and
Y.
Shibasaki
, in
2016 IEEE 66th Electronic Components and Technology Conference (ECTC)
(
IEEE
,
2016
), p.
159
.
34.
X.
Wei
,
S.
Kumagai
,
K.
Tsuzuku
,
A.
Yamamura
,
T.
Makita
,
M.
Sasaki
,
S.
Watanabe
, and
J.
Takeya
,
Flex. Print Electron.
5
,
015003
(
2020
).
35.
S.
Hong
,
S. P.
Park
,
Y.
Kim
,
B. H.
Kang
,
J. W.
Na
, and
H. J.
Kim
,
Sci. Rep.
7
,
16265
(
2017
).
36.
Y.-H.
Kim
,
J.-I.
Han
, and
S. K.
Park
,
IEEE Electron Device Lett.
33
,
50
(
2012
).
37.
J. K.
Jeong
,
H.
Won Yang
,
J. H.
Jeong
,
Y.-G.
Mo
, and
H. D.
Kim
,
Appl. Phys. Lett.
93
,
123508
(
2008
).